Verified project record

Tracer-Cloud/opensre

opensre is an open-source framework for building AI SRE agents that resolve production incidents, built to run on your own infrastructure. Users describe incidents in plain language and the agent gathers correlated logs, metrics, traces and recent deploys to answer with evidence-linked responses.

Project overview

It targets the real pain of incident evidence scattered across logs, metrics, traces, runbooks and Slack threads by correlating that evidence automatically, and it runs self-hosted with reversible identifier masking before external LLM calls and support for bringing your own model from multiple providers.

Project type: AI Agent · Infrastructure
Deployment: Refer to project documentation
License: Apache-2.0

Best for

AI engineers and operations teams who want a self-hosted framework for agents that resolve production incidents on their own infrastructure.

Key capabilities

Running `opensre` with no subcommand validates your account and starts a REPL where users describe incidents in plain language and use slash commands for session control, integrations and agent monitoring.
Automatically fetches relevant context and correlated logs, metrics, traces and recent deploys, reasons across connected systems, answers with evidence-linked responses, and suggests next steps and optional remediation actions.
Run one agent turn non-interactively with `opensre ask "why is checkout-api slow?"`, or drive the agent in-process from Python via `from bootstrap.embedded import start_embedded_session`.

Limitations and risks

The project is in Public Alpha: core workflows are usable for early exploration but not yet fully stable, and APIs and integrations may evolve.
Operation requires a bring-your-own model from an external LLM provider such as Anthropic, OpenAI, Ollama or Bedrock; inference is not bundled with the framework.

Getting started

Install with `curl -fsSL https://install.opensre.com | bash -s -- -gh`, run `opensre`, and sign in when prompted. Setup difficulty is documented as easy.

Evidence and sources

README: 🚧 Public Alpha: Core workflows are usable for early exploration, though not yet fully stable.
README: When something breaks in production, the evidence is scattered across logs, metrics, traces, runbooks, and Slack threads. OpenSRE is an open-source framework for AI SRE agents tha…
README: 🤖 **Full LLM flexibility** | Bring your own model — Anthropic, OpenAI, Codex, Ollama, Gemini, OpenRouter, TrustedRouter, NVIDIA NIM, Bedrock
README: **Interactive shell** — with no subcommand, `opensre` validates your account and starts a REPL (TTY required). You can exit and stay signed out, but the shell only opens for an ac…
README: 1. **Fetches** the relevant context and correlated logs, metrics, traces, and recent deploys 2. **Masks** sensitive identifiers (optional) before external LLM calls 3. **Reasons**…

Source updated at: 2026-10-12T20:21:09Z

Open original GitHub project